Segmentation methods of OCR systems in problems of automatic processing of archival documents

This paper describes the comparison of the modern optical character recognition systems aimed to find the systems, which do more precise segmentation, and to detect the capabilities of systems to allocate different types of areas. The results of the segmentation methods of OCR systems are analyzed. The effectiveness of the process of segmentation is evaluated. Based on the results of studies and observations made, recommendations to use for different types of documents are made.